When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I sent a brand new S2 fuel pump assembly to Deatschwerks for testing. Four tests were made. One pair of tests is run at 10 Volts to simulate low speed resistor mode and then also at 13.5 Volts to simulate high load conditions. Both conditions are tested on the factory Denso pump that comes in the OE housing and then also for the Deatschwerks DW65C 265LPH pump installed in an otherwise unmodified factory housing; four tests total.
The purpose of the tests is to determine the real output of the OE S2 pump assembly and then also the modified output from the 265LPH pump. This removes all the BS assumptions and guesstimations.
First and foremost observation is that the internal relief valve blows off around 50 psig and once that starts happening the true total output is unknown because what's going through the relief valve can't be measured without a lot of headaches and is just too much effort. If I had to do it again I probably would have asked for 45 psig to be used as a testing point, but I think we can't make some reasonable estimates there regardless. Ultimately what's going out the housing discharge fitting is all that matters to the engine.
The S2 pressure relief valve differs from the S1 in that it's an integral, fully enclosed, replaceable metal housing assembly that is held in place by the snap-lock for the sock filter. It can't blow apart and doesn't require melting, etc. modification like the S1 PRV. In theory you could swap it out for a different pressure set-point or adjustable PRV if you can find one with a similar body profile, or even make a plug to block it off entirely if you decide to install an external PRV instead. The overall design changes from the S1 pump are a definite improvement in many areas.
The OE Denso pump is pretty decent. At 13.5V and 45 psig it's output is approx. 210LPH. If fitted with a Kenne Bell 17V Boost-A-Pump it should be enough for high boost but not too crazy 2-rotor turbo engine on gasoline or a reasonable boost 2-rotor turbo engine on E-85. At the same volts and pressure the DW65C 265LPH pump has approx 230LPH output flow, so an improvement over OE, but maybe not really necessary for most. Assuming I ever get motivated and finish putting my RX8 with E85 NA engine back together it's certainly way more than enough for my needs. The 265LPH rating is for 40 psig and we can see it's actually exceeding this in the test. Of course everything is brand new, so no shocker. It appears that the OE housing and replaceable internal filter assembly when new isn't excessively restricted in any way. They also now offer a DW300 pump that will fit the same profile and is rated for 340LPH, but it hasn't been tested by me.
In the low-speed (10V) resistor mode, the two pumps are not too far apart at 45 psig; 145LPH for the OE Denso pump and 165 LPH for the DW65C pump. I haven't ever seen any testing for low speed mode and was just curious to have the data. Now we know.
The electrical amp data is self explanatory, but it provides a good idea of what electrical loads can be expected and planned for.
The tests cost me around $120 if I recall correctly, but then it came with the DW65C pump already installed in the OE housing. No biggie as I had already torn it down previously and will post up pictures of everything at some point. This is for the usual suspects that say I never contribute anything useful to the forum ... heh.
Some pictures of the pump module assembly internals and how (I believe) it functions. Special thanks to 9krpmrx8 for posting up some pics that I missed, which I included for additional clarification
New S2 pump assembly from Mazda
Disassembled ...
This is the basic pump and filter assemble. The The rough sock filter is at the bottom, the large white canister is the fine filter, and the pump is housed inside of it. So the pump draws in fuel from the sock filter, pressurizes it, discharges the fuel into the fine filter canister, and then the pressurized, filtered fuel is discharged into the white appendage on the bottom tight and out the black corrugated hose on the top to the engine.
At the bottom of the fuel filter discharge appendage is the metal relief valve. It discharges directly back into the fuel pump bowl. It's also removable/replaceable and secured in place by the sock filter attachment device. Once the sock filter assy. is removed, you can then remove the pump and the relief valve.
Here the sock filter is removed and the bottom of the pump and the relief valve can be grabbed onto and pulled from the fine filter canister. The fine filter canister is also replaceable from Mazda.
Some pics of the disassembled pump/filter assemble:
Top of the Denso pump.
Here you can see the Denso pump ID. This is a type "65C pump. Any 65C replacement/aftermarket pump should fit without modification. It's not interchangeable with an S1 pump. They're different lengths, diameters, etc. You have to use a "65C" design pump in the S2 module.
Fine filter cartridge with the pump and relief valve removed
Another view looking into the relief valve location
The removed sock rough filter and the relief valve. The relief valve has two O-ring seals. You can also see how the sock filter attachment also keeps the relief valve locked in place with attachment clips.
Getting into the nutty-gritty about the siphon assembly and pump module cup
So again, the is the fuel discharge appendage on the fine filter with the main fuel discharge hose out of the top and the pressure relief devise out the bottom. The small black line out the side is pressurized fuel that goes to the siphon device to pull fuel from the RH side of the saddle-style fuel tank over into the LH side.
Here you can see the pressurized fuel line to the siphon assembly. The black line going down the side f the module cut is the fuel being discharged from the siphon assembly down and into the bottom of the cup. I have the discharge end disconnected from the bottom cup inlet nipple for it.
Looking down into the pump module cup. You can see the cup inlet flapper valve (blue/orange device). The S2 flapper and inlet hole are larger than the S1 pump module. As the pump pulls fuel out of the cup the higher fuel level outside the pump module will push open the flapper valve and flow into to equalize it. You can also see the siphon device that has it's only siphon container housed within another smaller cup in the main fuel cup.
Here the siphon assembly has been removed from the siphon container. You can see the feed nipple at the bottom which the internal hose from the RH side of the fuel tank attaches to for pulling the fuel over into the pump module in the LH side. It appears to have an O-ring fit at the bottom, but I didn't try to remove it.
Another shot looking into the siphon container and side cup that it sits inside
And the siphon assembly itself. The black line from the bottom is the pressurized fuel supply from the pump/filter. It goes through the siphon and then out the other black line that attaches to the bottom of the main module cup. The siphon itself fits into the top of the siphon cup with an O-ring seal. As fuel flows through the siphon it pulls a vacuum on the siphon cup to pull fuel over from the RH side of the fuel tank.
Where the siphon discharge fuel attaches to the inlet nipple at the bottom of the module cup.
Thanks, that is some interesting info. One thing I don't understand is how the fuel line pressure can be higher than the relief valve set point. The spec is 54.4-60 psi, and on my car it is 59 psi at idle...
The unrestricted flow for 10V is 145LPH @ 45 psig. I could calculate it out, but the flow to the engine at idle is only a small fraction of that. Since the fuel system is non-return that means all the excess is going through the PRV, which is not that big. In short, dumping everything through the PRV creates a lot of backpressure on the pump at low flow to create that pressure spec. We didn't test that, but it could be done by throttling an outlet discharge valve back to a calculated idle flow rate.
EDIT: The issue with these charts and what you’ll see in actual practice in the tank is that Mazda way oversized the pump to crate very high bypass flow. Which when you restrict the the output flow to the engine at the much lower flow rate demand it has, the the restriction of the high bypass flow raises the pressure. So the general issue imo with many high flow scenarios is trying to maintain 60 psi pressure. Whereas most aftermarket systems with adjustable bypass regulators run around 42 - 45 psig, though typically with a boost signal to compensate fuel pressure higher relative to boost pressure.
In these charts they just didn’t restrict output flow down to where the OE engine operates. They can only measure output flow, not bypass flow. So from their perspective, once the internal bypass opens they can no longer accurately determine a true pump flowrate. They’re not looking at it from the perspective of the bypass being the siphon function to pull fuel from the opposite tank saddle. However, what you can derive from this is that if you’re tuning at 60 psig, but run the flow output to the engine higher that two things occur.
1. Bypass function decreases
2. Fuel pressure will drop.
The bigger issue is #2 (lol). You likely can’t adequately tune for the pressure drop with consistency. To me that creates a tuning issue, or potentially an engine goes boom issue. So in my mind, without an adjustable bypass and also possible eliminating the siphon by using a lower flow and pressure lift pump in the other saddle then potential issues may arise. Alternatively, a surge tank between the fuel tank and the engine such as the Radium FST-R with adjustable regulator that internally bypasses, is another way to tackle it, but still requires a basic understanding on how to set it up so that the pump module in the fuel tank still retains the siphon function properly by back-pressuring the output some.
Good write up, like posted before, i changed the oem pump inside s2 assembly for a denso 265 and i ran it like that for 1k miles last year on rew swap, no return.
This year ill install a external fpr and run a return line and hook it up to a radium jet venturi and see how it works. Im doing this because fuel pressures are unstable at the moment and will be hard to tune that way
The unrestricted flow for 10V is 145LPH @ 45 psig. I could calculate it out, but the flow to the engine at idle is only a small fraction of that. Since the fuel system is non-return that means all the excess is going through the PRV, which is not that big. In short, dumping everything through the PRV creates a lot of backpressure on the pump at low flow to create that pressure spec. We didn't test that, but it could be done by throttling an outlet discharge valve back to a calculated idle flow rate.
Interesting ..... I found when I put a DW300 in a S1 assy. that i had bad pressure spikes when the pump switched over to 'hi' . After drilling out the siphon , i lost the spikes , but pressure did not drop from where it was . I'm thinking maybe the PRV cracks at lower psi but doesn't full open till 55-60 .
Just looking at the way flow drops away more dramatically on the dw65 ......... maybe it's going to flow a lot more in the actual car as the flow increases to the injectors and the siphon isn't flowing much.
Wait a cotton pickin' minute ! is this the same 'Team' that disappeared without trace last year ?
Does anyone know if the S2 pump has the same pressure specs as the S1? I have an 04, which I have replaced the pump with an S2 pump and assembly as the S1 pump assembly gave me much trouble at the track (left-hand turns, etc. as discussed in many other threads). I bought a used S2 pump about 1.5 years ago (~30k miles at the time). I have put about 6k on it and done 10 track weekends on this time. Recently my car feels sluggish at high-end RPM +7k, my AFR are quite poor 13.3-13.6 at those RPM’s while data from last year was about 12.4. The Engine is quite recent (~1.5k), and I do feel the SSV opening for what I consider normal.
I have replaced coils, cables, and plugs. Checked for vacuum leaks and I haven’t found one, also cleaned the MAF and inspected the eccentric shaft position sensor (just in case). Current codes are P0410 (Air System), and P0037 (HO2S Heater Control Circuit Low Voltage), due to the removal of the air pump and the lack of a cat on the mid pipe.
My next step is to check the fuel pump. I installed the Aeromotive quick-connect and a ProSport fuel pressure gauge and I went through the procedure to test the pump as described in the service manual. The expected pressures are between 54-62psi, with a holding of +29psi. My current pressures are 52-55psi with holding pressure 48psi. I have not seen a dramatic decrease in pressure at high RPMs, which I was kind of hoping to blame the pump and get on with it, but it’s not the case. The running pressure does seem low, but I don’t know if the S2 pump could have different specs and if the test could be any different. I haven’t found any pressure test instructions specific for the S2 pump.
Does anyone have info on the S2 fuel pump pressures?
Does anyone know if the S2 pump has the same pressure specs as the S1? I have an 04, which I have replaced the pump with an S2 pump and assembly as the S1 pump assembly gave me much trouble at the track (left-hand turns, etc. as discussed in many other threads). I bought a used S2 pump about 1.5 years ago (~30k miles at the time). I have put about 6k on it and done 10 track weekends on this time. Recently my car feels sluggish at high-end RPM +7k, my AFR are quite poor 13.3-13.6 at those RPM’s while data from last year was about 12.4. The Engine is quite recent (~1.5k), and I do feel the SSV opening for what I consider normal.
I have replaced coils, cables, and plugs. Checked for vacuum leaks and I haven’t found one, also cleaned the MAF and inspected the eccentric shaft position sensor (just in case). Current codes are P0410 (Air System), and P0037 (HO2S Heater Control Circuit Low Voltage), due to the removal of the air pump and the lack of a cat on the mid pipe.
My next step is to check the fuel pump. I installed the Aeromotive quick-connect and a ProSport fuel pressure gauge and I went through the procedure to test the pump as described in the service manual. The expected pressures are between 54-62psi, with a holding of +29psi. My current pressures are 52-55psi with holding pressure 48psi. I have not seen a dramatic decrease in pressure at high RPMs, which I was kind of hoping to blame the pump and get on with it, but it’s not the case. The running pressure does seem low, but I don’t know if the S2 pump could have different specs and if the test could be any different. I haven’t found any pressure test instructions specific for the S2 pump.
Does anyone have info on the S2 fuel pump pressures?
I'm looking at the logs, but I see no short/long term fuel trims recorded. I use the dashcommand app; I may need to add the data channels. I will test it later this week and gather the data.
If you go to my original test data the relief valve on the S2 pump blows off earlier than the service hold pressure. In a dead head or low fuel flow scenario when the pump is running those pressures are obtainable because the relief valve can't flow enough to keep the pressure down. However, with the relief valve pressure being lower than the stated hold pressure I don't see how it could prevent it from blowing off until the pressure gets down to where the relief valve closes.
I'm looking at the logs, but I see no short/long term fuel trims recorded. I use the dashcommand app; I may need to add the data channels. I will test it later this week and gather the data.
I gathered some data, but it doesn't seem that the long term fuel trim data channel is recording properly, it is flat at zero the whole time while short term does fluctuate. I run for 15-20 minutes, but the long trim never changed. I may need to contact Palmer to check on the App settings/channels.
That remains to be seen. It works much better than the S1 pump module, has a higher flow rating and will support more HP for turbo/SC engines, doesn't fuel starve as easy as the S1 pump module in competition/racing high speed cornering events, and unlike the S1 pump is mostly serviceable for the internal filters and such which the S1 isn't. It's definitely an upgrade for the S1, but it requires a few mods to make the swap.
Here's the link to S1 DIY thread for making the swap
03-13-2016, 07:53 PM
